Cryptographic Scalability in the Smart Grid
Los Angeles Department of Power and Water
In the envisioned smart grid, massive numbers of computational devices will need to authenticate to each other. In the past, such technology would need to rest on a public key infrastructure (PKI) such as X.509. Today, many new cryptographic schemes are being proposed to solve the problem. However, deploying cryptography on such a large entity population—and doing the kinds of things we want the smart grid to do—raises many scalability challenges the community will need to address. Those challenges will only grow with the envisioned “Internet of Things.” The goal of this project was to develop high-fidelity, multi-scale models and use simulation to look for potential bottlenecks in this trust infrastructure. We recognized that on the transmission side, real-time is critical, and that the X.509 PKI standard didn’t work on the Border Gateway Protocol (BGP) with only 30k nodes, whereas the transmission side of the smart grid may have 100k nodes in the U.S. alone. At the same time, on the consumer side, revocation will be necessary, but it didn’t work with SSL servers, for which there are only 1 million correctly certified nodes worldwide (whereas there may be 1 billion consumer-side nodes in the U.S. smart grid, if we consider large appliances). Further, there may need to be attribute certificates; that has never been done before at the scale of the smart grid. On the modeling and simulation side, we saw the need for novel approaches to multi-scale modeling and simulation in order to capture dynamics of extremely large systems with sufficient fidelity. The initial stages of this project studied the extent to which the needed implementation would have to go beyond any current X.509 system in terms of size and functionality. We also looked at the hidden costs that might be involved with a much larger PKI, considering the smart grid’s needs and constraints. The modeling and simulation techniques we developed in this project will make it possible to quantify the costs of deploying PKI at scale in the smart grid and use the data to mitigate bottlenecks and other problems. Our approach will also extend to other large populations—such as the Internet of Things—requiring trust infrastructure.